algorithm theoretical baseline document (atbd) for product
TRANSCRIPT
Algorithm Theoretical Baseline
Document - ATBD-H50
(Product H50 – P-IN-LI)
Doc.No: SAF/HSAF/ATBD-50
Issue/Revision Index: 1.1
Date: 14/11/2019
Page: 1/13
Algorithm Theoretical Baseline Document (ATBD)
for product H50 – P-IN-LI
P-IN-LI Rainfall intensity from MTG-LI
Reference Number: SAF/HSAF/ATBD-50
Issue/Revision Index: 1.1
Last Change: 14/11/2019
Algorithm Theoretical Baseline
Document - ATBD-H50
(Product H50 – P-IN-LI)
Doc.No: SAF/HSAF/ATBD-50
Issue/Revision Index: 1.1
Date: 14/11/2019
Page: 2/13
DOCUMENT CHANGE RECORD
Issue / Revision Date Description
0.1 12/01/2017 Preliminary version prepared for PDR
1.0 01/06/2018 Version prepared for MTG CDR
1.1 14/11/2019 Version prepared for MTG SIRR
Index List of figures ........................................................................................................................................ 2
1. Introduction .................................................................................................................................. 3
1.1. The MTG program ................................................................................................................. 4
1.2. The Flexible Combined Imager (FCI) .................................................................................. 4
1.3. Lightning Imager ................................................................................................................... 5
1.4. P-IN-LI Requirements ........................................................................................................... 5
2. Algorithm description .................................................................................................................. 5
3. Operational chain ......................................................................................................................... 6
4. P-IN-LI – SEVIRI based prototype ........................................................................................... 7
5. References ..................................................................................................................................... 9
Annex 1: Acronyms ............................................................................................................................ 10
List of figures Fig 1: P-IN-MHS precipitation retrieval by microwave sensor on polar satellite (turquoise elliptical footprint) and radar (with colour scale). .............................................................................................. 3
Fig 2: P-IN-MHS ellipsis as above and occurred lightnings in 15 minutes centred in microwave observation timestamp. ........................................................................................................................ 4
Fig 3: Histogram of IR brightness temperatures at lightning locations ............................................... 6
Fig 4: H50 prototype rainfall retrieval and radar rain rate .................................................................. 8
Algorithm Theoretical Baseline
Document - ATBD-H50
(Product H50 – P-IN-LI)
Doc.No: SAF/HSAF/ATBD-50
Issue/Revision Index: 1.1
Date: 14/11/2019
Page: 3/13
1. Introduction One of the objectives of the “EUMETSAT Satellite Application Facility on Support to Operational
Hydrology and Water Management (H-SAF)” is to provide new SAF satellite-derived products from
existing and future satellites with sufficient time and space resolution to satisfy the needs of operational
hydrology. Active and passive microwave sensors aboard polar-orbiting satellites provide observations
physically connected to precipitation but at a temporal resolution (once or twice a day) that
considerably limits their utility in many applications.
It's even possible that a severe storm can start and finish between two passages of satellites. In the Fig
1 there is a comparison between precipitation retrieval by microwave sensor on polar satellite
(turquoise elliptical footprint) and radar (with color scale). Precipitation volume estimation between
satellite and radar is very similar but the field shape is different, in detail MW space measurement
could not distinguish inside active pixel.
Fig 1: P-IN-MHS precipitation retrieval by microwave sensor on polar satellite (turquoise elliptical footprint) and
radar (with colour scale).
In these conditions, the cloud-top infrared (IR) observations from geostationary satellites remain the
main data source for near-continuous monitoring from space. The main difficulty in IR-based rainfall
estimation is the fact that precipitation can only be inferred from cloud-top observations. Because a
cloud’s brightness temperature, or other cloud-top characteristics, may correspond to different surface
rainfall rates depending on the rain regime or other factors, IR-based rainfall estimation may be
improved significantly only by considering additional information. Past studies (Adler and Negri 1988;
Anagnostou et al. 1999) have shown that minima in the IR temperature array may be associated with
enhanced convective regions in the cloud, and substantially better estimates may be obtained by using
distinct algorithms for convective and stratiform precipitation. The ability to detect convective cores
and rain/no-rain boundaries, however, is limited from only being able to ‘‘see’’ the features at cloud
top. Thus, the quantification of the convective rainfall is subject to significant uncertainty, despite the
fine time resolution of the data. To reduce this uncertainty, additional information should be
considered. Lightning data may represent such information; lightning data contain useful information
for rainfall estimation, in particular in describing the field shape and the location of higher intensity
Algorithm Theoretical Baseline
Document - ATBD-H50
(Product H50 – P-IN-LI)
Doc.No: SAF/HSAF/ATBD-50
Issue/Revision Index: 1.1
Date: 14/11/2019
Page: 4/13
rain cluster (De Leonibus et al. 2008). In the Fig 2 there is the electrical activity in the same area. The
position of convective precipitation is almost entirely seen by means of lightning rate and locations.
Fig 2: P-IN-MHS ellipsis as above and occurred lightnings in 15 minutes centred in microwave observation
timestamp.
1.1. The MTG program The MTG program provides meteorological imagery over Europe and Africa and maintains continuity
of the Meteosat program, continuing and expanding the service provide by MSG.
The MTG system will be operated by EUMETSAT and will provide Europe’s national meteorological
services as well as international users and science community with improved imaging and new infrared
sounding capabilities for both meteorological and climate applications. The MTG program is
established as a common undertaking between EUMETSAT and the European Space Agency (ESA).
ESA is responsible for the development of the MTG space segment. EUMETSAT is responsible for
the development of the MTG ground segment and service provision. Thanks to advances in technology,
MTG will also provide an enhanced service compared to the current MSG system, contributing
significant improvements to the existing service with an improved imagery mission (Flexible
Combined Imager - FCI) and introducing new sounding and lightning missions (Lightning Imager -
LI) from a geostationary orbit.
1.2. The Flexible Combined Imager (FCI) The FCI will provide follow-on services to the Full Disc Scanning Service (FDSS) and Rapid Scanning
Service (RSS) currently provided by the Meteosat Second Generation (MSG) Spinning Enhanced
Algorithm Theoretical Baseline
Document - ATBD-H50
(Product H50 – P-IN-LI)
Doc.No: SAF/HSAF/ATBD-50
Issue/Revision Index: 1.1
Date: 14/11/2019
Page: 5/13
Visible and Infrared Imager (SEVIRI). The FCI has channels over 16 spectral ranges covering visible
to infrared wavelengths.
1.3. Lightning Imager
The Lightning Imager (LI) will offer improvements for nowcasting by delivering information on total
lightning (Intra Cloud (IC) and Cloud to Ground (CG)). The instrument will bring full near real-time
total lightning detection capabilities.
The benefit of the LI mission is that it will observe total lightning continuously and simultaneously
over the full disk, providing the information to the users with an extremely high timeliness.
The LI mission will be able to detect, monitor, track and extrapolate, in time, the development of active
convective areas and storm life cycles — critical for nowcasting and very short range forecasting of
severe weather events.
Additionally, one method of assessing the impact of climate change on thunderstorm activity is to
globally monitor and long-term analyze the lightning characteristics, which would require a long-term
stable and spatially homogeneous lightning observing system.
Lightning is also a major source of harmful nitrogen oxides (NOx) in the atmosphere. NOx plays a key
role in the ozone conversion process and acid rain generation. A detailed knowledge of the global
distribution of the total lightning (CG + IC) is a prerequisite for studying and monitoring the physical
and chemical processes in the atmosphere, regarding NOx.
The LI on MTG will complement the two NOAA GLMs (Geostationary Lightning Mapper) on the
GOES-R and the GOES-S satellites, thus contributing, in the long term, to near global coverage.
1.4. P-IN-LI Requirements For detailed description of requirements characteristics see Product Requirement Document
2. Algorithm description According to Grecu et al. (2000), to determine a quantitative relationship for rainfall estimation using
lightning and Infrared data a bivariate linear regression for clusters rain volume has to be used:
𝑅𝑅 = (𝑏0 +𝑏1 ∗ 𝑆
𝑁+ 𝑏2 ∗ 𝑇) ∗ 𝑁
where RR is the radar measured cluster rain volume in on hour (mm*h-1*km2), S is the number of
strokes in the cluster, T (°K) is the minimum in the 10.8μm brightness temperature in the cluster and
N is the number of pixels in the cluster. The coefficients b0, b1, and b2 has to be calculated by means
of least squares regression.
To disaggregate cluster rain volume to its individual pixel, it has to build the histogram counting all
pixels associated with lightning that exist in the dataset for a given temperature. The probability
P(T) = P(that raining pixel temperature > T)
Algorithm Theoretical Baseline
Document - ATBD-H50
(Product H50 – P-IN-LI)
Doc.No: SAF/HSAF/ATBD-50
Issue/Revision Index: 1.1
Date: 14/11/2019
Page: 6/13
is retrieved from the histogram features. The disaggregating procedure uses these probabilities to
compute weights for each individual pixel of a convective cluster. A pixel weight is defined as the ratio
of the pixel’s probability to the sum of probabilities of all pixels located within the same convective
cluster.
An example of such a histogram is in Fig 3. It has been built using a dataset composed by 27 days of
10.8 μm SEVIRI images and lightning data provided by LAMPINET, lightning network of Italian Air
force meteorological service.
Fig 3: Histogram of IR brightness temperatures at lightning locations
3. Operational chain The proposed algorithm combining brightness temperature and lightning for the estimation of
convective rainfall can be summarized as follows:
1. load Flexible Combined Imager (FCI) IR data and compute 10.5μm brightness
temperature;
2. compute lightning number from Lightning Imager (LI) within a 10 minutes’ window,
centered around IR sampling time;
3. individuate lightning clusters;
4. estimate rain volume of each cluster with the equation above;
5. allocate the rain volume of each cluster to pixels based on probability of the brightness
temperature histogram.
Algorithm Theoretical Baseline
Document - ATBD-H50
(Product H50 – P-IN-LI)
Doc.No: SAF/HSAF/ATBD-50
Issue/Revision Index: 1.1
Date: 14/11/2019
Page: 7/13
The histogram of IR brightness temperature should be different accordingly the season and
climatological area (mid-latitudes, tropics, etc.).
4. P-IN-LI – SEVIRI based prototype
To verify the effectiveness of the algorithm, a prototype using SEVIRI images and lightning data from
LAMPINET has been realized.
A dataset composed by 27 days of gathered data (radar rainfall rate, 10.8μm SEVIRI images, lightning
data, P-IN-MHS precipitation rate at ground) from 3rd March 2011 to 30th June 2011 is being used.
Lightning data are provided by LAMPINET, lightning network of Italian Air force meteorological
service. The network is based on Vaisala technology with 15 IMPACT ESP sensors distributed on the
Italian peninsula and islands. Performances of the network can reach a detection efficiency of 90% and
location accuracy of 0.5 km all over Italian area. Brightness temperature of SEVIRI channel 9 (10.8μm)
is processed straight from HRIT raw files. P-IN-MHS is based on the instruments AMSU-A and MHS
flown on NOAA and MetOp satellites. These cross-track scanners provide images with constant
angular sampling across Track.
Radar rainfall rate is provided by DPC (Dipartimento di Protezione Civile, Italian Civil Protection
Department). Currently, the National Radar Network is composed of 21 radars that work in the C band.
The volume made available from each site, with a generation frequency of at least 10 minutes, is pre-
processed according to a set of techniques to de-clutter and resampled at a resolution of 1km. The
product of SRI calculates the precipitation to the ground by applying an algorithm on volumetric data
of the PPI reflectivity at lower elevation between those acquired which meet the quality criteria in the
planning stage. The reflectivity values are converted to measure precipitation (rain rate mm/h)
according to the Marshall Palmer equation.
In each day of the dataset there are multiple storms. Dataset contains 4703 lightning's clusters. This
dataset has been used for the calibration of the prototype (i.e. least squares regression to determinate
the coefficients b0, b1, and b2). The values of the parameters identified by this calibration are b0 =
0.5131, b1 = 0.2373 and b2 = -0.0014.
To characterize statistically the agreement between radar rainfall rate (considered as “true” value) and
lightning observations three performance scores has been used: the probability of detection (POD), the
false alarm rate (FAR), and the critical success index (CSI), defined as
𝑃𝑂𝐷 =𝑛𝑠𝑢𝑐𝑐𝑒𝑠𝑠
𝑛𝑠𝑢𝑐𝑐𝑒𝑠𝑠 + 𝑛𝑓𝑎𝑖𝑙𝑢𝑟𝑒
𝐹𝐴𝑅 =𝑛𝑓𝑎𝑙𝑠𝑒𝑎𝑙𝑎𝑟𝑚
𝑛𝑠𝑢𝑐𝑐𝑒𝑠𝑠 + 𝑛𝑓𝑎𝑙𝑠𝑒𝑎𝑙𝑎𝑟𝑚
𝐶𝑆𝐼 =𝑛𝑠𝑢𝑐𝑐𝑒𝑠𝑠
𝑛𝑠𝑢𝑐𝑐𝑒𝑠𝑠 + 𝑛𝑓𝑎𝑙𝑠𝑒𝑎𝑙𝑎𝑟𝑚 + 𝑛𝑓𝑎𝑖𝑙𝑢𝑟𝑒
where nsuccess, nfailure and nfalsealarm are the number of successes, failures, and false alarms, respectively,
in the comparison. A comparison yields a success when there is lightning and a radar rainfall rate
Algorithm Theoretical Baseline
Document - ATBD-H50
(Product H50 – P-IN-LI)
Doc.No: SAF/HSAF/ATBD-50
Issue/Revision Index: 1.1
Date: 14/11/2019
Page: 8/13
greater than 10 mm/h, a false alarm when there is lightning but not rainfall, and a failure when there is
rainfall greater than 10 mm/h with no lightning indications.
The software has been tested on 873 lightning's clusters and the results are:
POD = 0.48
FAR = 0.34
CSI = 0.30
In Fig 4 there is a visual comparison in a case study (7th July 2011 02:15 UTC) between H50 prototype
OUTPUT and rainfall rate from radar.
Fig 4: H50 prototype rainfall retrieval and radar rain rate
The general conclusion of this study is that lightning data contain useful information for satellite
rainfall estimation, mainly in convective phenomena where MW retrieval can presents some problems.
Good scores of P-IN-LI prototype suggests a possible use as product in HSAF project to improve
rainfall estimation of MW and IR techniques. In future the prototype will be tested with LI and after a
calibration period it will be able to work with the sensor data.
Algorithm Theoretical Baseline
Document - ATBD-H50
(Product H50 – P-IN-LI)
Doc.No: SAF/HSAF/ATBD-50
Issue/Revision Index: 1.1
Date: 14/11/2019
Page: 9/13
5. References
Grecu, Mircea, Emmanouil N. Anagnostou, and Robert F. Adler. "Assessment of the use of lightning
information in satellite infrared rainfall estimation." Journal of Hydrometeorology 1.3 (2000): 211-
221.
Adler, Robert F., and Andrew J. Negri. "A satellite infrared technique to estimate tropical convective
and stratiform rainfall." Journal of Applied Meteorology 27.1 (1988): 30-51.
Anagnostou, Emmanouil. "A satellite infrared technique for diurnal rainfall variability studies."
(1999).
De Leonibus, Luigi, et al. "Rainfall Field Reconstruction over Italy Through LAMPINET Lightning
Data" (2008).
Algorithm Theoretical Baseline
Document - ATBD-H50
(Product H50 – P-IN-LI)
Doc.No: SAF/HSAF/ATBD-50
Issue/Revision Index: 1.1
Date: 14/11/2019
Page: 10/13
Annex 1: Acronyms AMSR2 Advanced Microwave Scanning Radiometer 2 AMSU Advanced Microwave Sounding Unit (on NOAA and MetOp) ATDD Algorithms Theoretical Definition Document ATMS Advanced Technology Microwave Sounder AU Anadolu University (in Turkey) BfG Bundesanstalt für Gewässerkunde (in Germany) CAF Central Application Facility (of EUMETSAT) CDOP Continuous Development-Operation Phase CESBIO Centre d'Etudes Spatiales de la BIOsphere (of CNRS, in France) CGMS Coordination Group for Meteorological Satellites CMAP Climate Prediction Center Merged Analysis of Precipitation CM-SAF SAF on Climate Monitoring COMet Centro Operativo per la Meteorologia (in Italy) CNR Consiglio Nazionale delle Ricerche (of Italy) CNRS Centre Nationale de la Recherche Scientifique (of France) COSMO-ME Consortium for Small-Scale Modelling - version for Mediterranean DMSP Defence Meteorological Satellite Program DPC Dipartimento Protezione Civile (of Italy) EARS EUMETSAT Advanced Retransmission Service ECMWF European Centre for Medium-range Weather Forecasts EDC EUMETSAT Data Centre, previously known as U-MARF EUM Short for EUMETSAT EUMETCast EUMETSAT’s Broadcast System for Environmental Data EUMETSAT European Organisation for the Exploitation of Meteorological Satellites FMI Finnish Meteorological Institute FTP File Transfer Protocol GEO Geostationary Earth Orbit GPCP Global Precipitation Climatology Project GPI GOES Precipitation Index GPM Global Precipitation Measurement GRAS-SAF SAF on GRAS Meteorology H-SAF SAF on Support to Operational Hydrology and Water Management IMWM Institute of Meteorology and Water Management (in Poland) IPF Institut für Photogrammetrie und Fernerkundung (of TU-Wien, in Austria) IPWG International Precipitation Working Group IR Infra Red IRM Institut Royal Météorologique (of Belgium) (alternative of RMI) ISAC Istituto di Scienze dell’Atmosfera e del Clima (of CNR, Italy) ITU İstanbul Technical University (in Turkey) LATMOS Laboratoire Atmosphères, Milieux, Observations Spatiales (of CNRS, in France) LEO Low Earth Orbit LHN Latent Heat Nudging LSA-SAF SAF on Land Surface Analysis Météo France National Meteorological Service of France METU Middle East Technical University (in Turkey) MHS Microwave Humidity Sounder (on NOAA 18 and 19, and on MetOp) MW Micro Wave NMA National Meteorological Administration (of Romania) NOAA National Oceanic and Atmospheric Administration (Agency and satellite) NWC-SAF SAF in support to Nowcasting & Very Short Range Forecasting NWP Numerical Weather Prediction NWP-SAF SAF on Numerical Weather Prediction
Algorithm Theoretical Baseline
Document - ATBD-H50
(Product H50 – P-IN-LI)
Doc.No: SAF/HSAF/ATBD-50
Issue/Revision Index: 1.1
Date: 14/11/2019
Page: 11/13
O3M-SAF SAF on Ozone and Atmospheric Chemistry Monitoring OMSZ Hungarian Meteorological Service OSI-SAF SAF on Ocean and Sea Ice PMW Passive Micro-Wave PP Project Plan PUM Product User Manual PVR Product Validation Report QPF Quantitative Precipitation Forecast REMET Reparto di Meteorologia (in Italy) RMI Royal Meteorological Institute (of Belgium) (alternative of IRM) SAF Satellite Application Facility SEVIRI Spinning Enhanced Visible and Infra-Red Imager (on Meteosat from 8 onwards) SHMÚ Slovak Hydro-Meteorological Institute SSM/I Special Sensor Microwave / Imager (on DMSP up to F-15) SSMIS Special Sensor Microwave Imager/Sounder (on DMSP starting with S-16) STD Standard Deviation SYKE Suomen ympäristökeskus (Finnish Environment Institute) TKK Teknillinen korkeakoulu (Helsinki University of Technology) TSMS Turkish State Meteorological Service TU-Wien Technische Universität Wien (in Austria) U-MARF Unified Meteorological Archive and Retrieval Facility UniFe University of Ferrara (in Italy) URD User Requirements Document UTC Universal Coordinated Time VIS Visible WMO World Meteorological Organization ZAMG Zentralanstalt für Meteorologie und Geodynamik (of Austria)